Summary Background : We evaluated the relationship between Helicobacter pylori and various factors associated with gastric cancer in two areas in Japan with different risks for mortality due to gastric cancer. Methods : A total of 250 sera from Niigata and 209 from Okinawa were used. H. pylori antibody and CagA antibody were measured by antigen‐specific ELISAs. Serum gastrin and pepsinogen levels were determined by RIA. Results : Although there was no significant difference in H. pylori prevalence among the persons in Niigata (50%) and Okinawa (42%), CagA prevalence in these populations was significantly different, at 41% and 26%, respectively (OR = 1.98, 95%CI: 1.33–2.95, P < 0.01). Serum gastrin levels in Niigata were significantly lower than those in Okinawa in H. pylori‐negative persons (P < 0.01). The serum pepsinogen I/II ratio in Niigata was significantly lower than that in Okinawa in H. pylori positive persons (P < 0.01), whereas there was no significant difference in H. pylori‐negative persons. Among those positive for H. pylori, serum pepsinogen I/II ratio in Niigata was significantly lower than that in Okinawa in CagA‐negative persons (P < 0.01), whereas no significant difference was observed in CagA‐positive persons. Conclusions : These results suggest that the difference in the mortality ratio of gastric cancer between Niigata and Okinawa is mainly associated with the difference between areas in the prevalence of cagA‐positive strains rather than that of H. pylori itself.